SAT Scores (Normalized) - And Applications for Haddonfield

In previous blog posts, I talked about school district comparables, taxes in New Jersey -- and Haddonfield school enrollment.  My goal with most issues is to "frame" things -- to get an understanding of how things work and where things stand. 
  
The College Board (which administers the SAT) publishes various SAT raw data, but it is hard to tell much from average SAT scores, by state.  Different regions of the country have different demographics and different goals.  For example, in New Jersey and New York, roughly 80% of students take the SAT.  However, in many Midwest states such as Iowa, Kentucky, and Arkansas, only about 5% of the students take the SAT.  

If only the top students in a given state take the SAT, average test scores are skewed (upwards).  On the other hand, some states have the lofty goal of obtaining very high SAT participation rates.  Over recent years, Delaware has had 100% SAT participation while Maine has 94% participation.  Some students are not interested in college or the SAT, so the results for these states are skewed downward.   

Based on the distribution of SAT results over the past few years, I normalized test results, taking into account SAT scores from various states AND the participation rates in each state.  In a nutshell, the goal is to equalize and adjust results for various participation rates.  Happily for New Jersey residents, our state places in the top quartile.  Over the past few years, the average SAT score in New Jersey is 1516 (with 78% participation). Other states, known for good schools -- such as Massachusetts and Connecticut -- also place in the top quartile.  Note that I purposely display the results only in quartiles because I feel like the world is sometimes too competitive -- when our goal is mainly to use research to improve society.

Please click on this link for the SAT Preparedness, by state.  

New Jersey comes out in the top quartile, so our state does a good job in terms of education and SAT preparedness.  

Applications for Haddonfield

This kind of approach (thoughtful analysis, improved comparisons, and in this case: normalizing data for participation rates) can also be applied to our school district's data to better understand strengths and weaknesses. We may be able to discover meaningful trends and help improve and/or better understand our results.  

More specifically, in this case, Haddonfield recently compared our students' SAT scores versus those of other "J Districts" across the state.  Our "average SAT scores" are decent -- but actually appear lower due to the good fact that we have very high participation rates!  
Adjusting our results based on our 100% participation -- or alternatively, adjusting other J districts for participation as low as the 70%-80% range -- would improve comparisons and put things on a more even playing field.  Several people I discussed this with -- including some involved with our schools -- are interested in this type of work.  

I have worked with data from the College Board as well as Great Schools.  In my work with college admissions folks, we are often trying to "normalize" data; that is: discover which students have "done what" -- with "what they have..."  It is noteworthy that SAT scores are only one measure of academic success.  However, Haddonfield also does well with proficiency in other standardized tests, and other measures of serving our students.  

As I study various topics, I am happy to see that we, in Haddonfield, are generally doing well in terms of usage of tax dollars, tax rate comparisons to peers, and many academic indicators.  However, we ARE getting squeezed with rising costs and limited budgets -- and I think I can be useful for the BOE with planning for the future, budgeting, forecasting -- and thoughtful analysis.    
Thank you, 
Carlton Chin
http://vote4chin.com/